The OpenAI Economic Research Team published an analysis showing that ChatGPT adoption broadened in the first quarter of 2026. The report finds growth across age groups, a rising share of users with typically feminine names, and relative increases in country-level usage.
Demographics and usage patterns
According to the report, users with typically feminine names now account for over half of users for whom gender can be inferred, and their share increased in early 2026 after reaching approximate parity the previous year. Messages from all age groups rose with overall growth; users under 35 still produced the largest share of messages, but the share from users over 35 increased in Q1.
Geography and workplace activity
OpenAI ranked countries by messages sent per capita to assess relative usage. The ten fastest-rising countries in per-capita rank between 2025 Q4 and 2026 Q1 were:
- Dominican Republic: 53 → 44 (+9)
- Haiti: 91 → 82 (+9)
- Japan: 43 → 35 (+8)
- Mexico: 60 → 54 (+6)
- Tanzania: 102 → 96 (+6)
- Brazil: 47 → 42 (+5)
- Costa Rica: 38 → 33 (+5)
- Myanmar: 99 → 94 (+5)
- Papua New Guinea: 109 → 104 (+5)
- Austria: 15 → 11 (+4)
The analysis covers messages on ChatGPT consumer plans (Free, Go, Plus, and Pro) and explicitly excludes Codex as well as ChatGPT Enterprise and ChatGPT for Education, so it likely understates total workplace and educational use.
Within work-related activity on personal accounts, creating written and visual materials continued to lead but declined over time as more specialized tasks grew. The fastest-growing workplace tasks cited were content creation, health-related documentation, and information retrieval. The report notes that work-related use on consumer accounts remained consistent with past trends and that users appeared to adopt repeatable use cases.
OpenAI Signals describes this dataset as a continuing resource for researchers and policymakers; the report says usage data for research and enterprise studies is available for download through OpenAI Signals’ data releases.
Original source: OpenAI News